jina.hubble.helper module#
Module for helper functions for Hub API.
- jina.hubble.helper.get_hub_packages_dir()[source]#
Get the path of folder where the hub packages are stored
- Return type
Path
- Returns
the path of folder where the hub packages are stored
- jina.hubble.helper.get_cache_db()[source]#
Get the path of cache db of hub Executors
- Return type
Path
- Returns
the path of cache db of hub Executors
- jina.hubble.helper.get_download_cache_dir()[source]#
Get the path of cache folder where the downloading cache is stored
- Return type
Path
- Returns
the path of cache folder where the downloading cache is stored
- jina.hubble.helper.get_hubble_url_v1()[source]#
Get v1 Hubble Url
- Return type
str
- Returns
v1 Hubble url
- jina.hubble.helper.get_hubble_url_v2()[source]#
Get v2 Hubble Url
- Return type
str
- Returns
v2 Hubble url
- jina.hubble.helper.parse_hub_uri(uri_path)[source]#
Parse the uri of the Jina Hub executor.
- Parameters
uri_path (
str
) – the uri of Jina Hub executor- Return type
Tuple
[str
,str
,str
,str
]- Returns
a tuple of schema, id, tag, and secret
- jina.hubble.helper.is_valid_huburi(uri)[source]#
Return True if it is a valid Hubble URI
- Parameters
uri (
str
) – the uri to test- Return type
bool
- Returns
True or False
- jina.hubble.helper.md5file(file_path)[source]#
Retrun the MD5 checksum of the file
- Parameters
file_path (
Path
) – the file to check md5sum- Return type
str
- Returns
the MD5 checksum
- jina.hubble.helper.unpack_package(filepath, target_dir)[source]#
Unpack the file to the target_dir.
- Parameters
filepath (
Path
) – the path of given filetarget_dir (
Path
) – the path of target folder
- jina.hubble.helper.archive_package(package_folder)[source]#
Archives the given folder in zip format and return a data stream. :type package_folder:
Path
:param package_folder: the folder path of the package :rtype:BytesIO
:return: the data stream of zip content
- jina.hubble.helper.download_with_resume(url, target_dir, filename=None, md5sum=None)[source]#
Download file from url to target_dir, and check md5sum. Performs a HTTP(S) download that can be restarted if prematurely terminated. The HTTP server must support byte ranges.
- Parameters
url (
str
) – the URL to downloadtarget_dir (
Path
) – the target path for the filefilename (
Optional
[str
]) – the filename of the downloaded filemd5sum (
Optional
[str
]) – the MD5 checksum to match
- Return type
Path
- Returns
the filepath of the downloaded file
- jina.hubble.helper.upload_file(url, file_name, buffer_data, dict_data, headers, stream=False, method='post')[source]#
Upload file to target url
- Parameters
url (
str
) – target urlfile_name (
str
) – the file namebuffer_data (
bytes
) – the data to uploaddict_data (
Dict
) – the dict-style data to uploadheaders (
Dict
) – the request headerstream (
bool
) – receive stream responsemethod (
str
) – the request method
- Returns
the response of request
- jina.hubble.helper.disk_cache_offline(cache_file='disk_cache.db', message='Calling {func_name} failed, using cached results')[source]#
Decorator which caches a function in disk and uses cache when a urllib.error.URLError exception is raised If the function was called with a kwarg force=True, then this decorator will always attempt to call it, otherwise, will always default to local cache.
- Parameters
cache_file (
str
) – the cache filemessage (
str
) – the warning message shown when defaulting to cache. Use “{func_name}” if you want to print the function name
- Returns
function decorator
- jina.hubble.helper.is_requirements_installed(requirements_file, show_warning=False)[source]#
Return True if requirements.txt is installed locally :type requirements_file:
Path
:param requirements_file: the requirements.txt file :type show_warning:bool
:param show_warning: if to show a warning when a dependency is not satisfied :rtype:bool
:return: True or False if not satisfied